基于IDBO-CNN-BiLSTM锂电池剩余使用寿命预测OA

The SOH(State of Health)and RUL(Remaining Useful Life)of batteries are important evaluation in-dicators for battery health management.In view of the problem that it is difficult to accurately predict the remaining use-ful life of lithium batteries due to the influence of many complex factors during their use,this study proposes a hybrid prediction model based on IDBO-CNN-BiLSTM(Improved Dung Beetle Optimizer-Convolutional Neural Networks-Bi-directional Long Short-Term Memory).By analyzing the state of the lithium battery during the charging process,nine HF(Health Factor)are extracted.The strongly correlated health factors are screened out through the Pearson cor-relation coefficient and used as the input of the model.The chaotic initialization Tent mapping is adopted to generate the initial positions of the dung beetles,and the sine-cosine strategy is used to optimize the positions of the stealing dung beetles.This solves the local convergence problem caused by the initialization of the DBO(Dung Beetle Optimi-zer)algorithm and optimizes the balance of the DBO algorithm,improving the stability of the prediction.Experiments are carried out based on the publicly available lithium battery aging dataset provided by NASA(National Aeronautics and Space Administration),and different models are used to predict the SOH of NASA's lithium batteries.The results show that the proposed method has a smaller error and has certain application value.